Concentric Tourist Spatial Model and Extraction of Relational Distance

Abstract

Since the second half of the 1970's, there have been many studies in the literature of tourism on the spatial behavior pattern. But most of these studies have failed to built a general model of tourist space, focusing too closely on specific facts derived from simple correlation analysis between a certain spatial pattern and other socioeconomic or spatio-temporal variables. In the present article, we discuss the fundamental geographical notions of region and distance, creating a concentric tourist spatial model by integrating two existing models: S. C. Plog's (1973) and J.-M. Miossec's (1977). What we mean by the concentric model is that the trip “distance” influences the tourist's behavior, perception and the frequency of trips, and vice versa.We collected the data for this study in September 1992 at an elementary school in Yokohama (a large city situated about 30km south of Tokyo). It is a binary matrix of 338 recreational family trips from Yokohama in the line and 94 tourist-/trip-variables in the row which indicate different types of tourists and trip patterns. Trip-variables include destination (s), trip duration, trip organizer, group composition, means of transportation, nature of activity at the destination, tourist's image of the destination, etc. Tourist-variables include quantity of information and ability to use it, duration of holidays, expenditures for tourism and frequency of recreational trips per year, etc.First of all, we showed how several “tourism regions” (S. L. J. Smith, 1989) are created by the two “regionalizations” (Smith): the regionalization based on the denomination by tourists of the destination zone and the regionalization based on the route taken in a real trip.Next, we summarized the trip matrix data and verified the concentric spatial model, applying Hayashi's “theory of the quantification 3” -mathematically almost the same analysis as the correspondence analysis in the Anglo-Saxon or French world. This analysis of quantification changes the original and arbitrary arrangement of the trips on the axis of the line into a new significant and revealing arrangement (simultaneously, every trip is given a quantity indicating its point on the axis). In other words, this method locates closely on the axis trips which resemble each other in their reaction pattern to the variables, and it distances trips which do not resemble each other. This analysis also changes the original arrangement of the variables on the axis of the row into a new meaningful arrangement; it locates closely/distances the variables according to their reaction pattern to the trips. This analysis replaces, in fact, both the trips and the variables on the axes in order to keep the highest possible correlation between the quantified trips and the quantified variables. The highest possible correlation means the new arrangements of the trips and of the variables have the same structure and the same meaning. Thus this analysis creates from one original binary matrix several new binary matrices independent from each other which summarize the data structure of tourist behavior, whereby the first matrix is the most important, and the second matrix is the second most important. We can explain these new matrices with the variable-axes.We found three important variable-axes: the trip “distance” (the most important element), the nature of the activity in the destination zone and the tourist's ability to travel (economic status and information level). Observing the three-dimensional data space composed by these axes, we found one dominant and five secondary variable masses of the tourist's profile and behavior pattern. The dominant one is about a two day 100km family trip, not planned by a tourist agency but by the family itself, for the purpose of enjoying nature and practicing sports.